Trending AI projects

MLflow is an Apache‑2.0‑licensed open‑source platform for managing the full lifecycle of machine‑learning and generative‑AI projects. It provides experiment tracking, model registry, deployment tools, and a new GenAI stack (tracing, evaluation, prompt registry, optimization, and an AI gateway) with integrations for dozens of LLM providers and agent frameworks.

gpt‑with‑taste is a repository that provides a single, author‑maintained system prompt (“recipe”) which can be compiled into native instructions for several LLM coding assistants (Codex, Claude, Kiki, Pi, DeepSeek, etc.). The recipe adds judgment, execution discipline, collaboration handling, and visual‑craft focus to model replies. Installation is a one‑line Node script that drops prompt files (and optional hooks) into the user’s Codex configuration, after which the model produces more concise, purposeful outputs, as shown in side‑by‑side case studies.

An AI-powered video note-taking tool that converts videos from Bilibili, YouTube, and other platforms into structured Markdown notes with automatic summaries and screenshots.

Chatbox is a cross-platform desktop and mobile client that provides a unified interface for popular AI models like ChatGPT, Claude, and local LLMs. It simplifies AI access by offering a local-first, easy-to-install application with features for prompt management and team collaboration.

ClawBench is an open‑source benchmark that evaluates AI browser agents on real‑world online tasks (travel booking, food ordering, job applications, etc.). It ships two corpora of 281 tasks across 163 live sites, records each run with a 5‑layer trace, and scores outcomes with an LLM‑based evaluator. Install via `pip install clawbench-eval` (or `uv tool install`), configure a model judge, and run tasks inside Docker/Podman using the `clawbench` CLI. Results appear on a HuggingFace leaderboard, and the framework is extensible for new tasks or agents.

An AI-powered low-code platform for Java enterprises that generates full-stack systems from natural language and provides a RAG-based AI application builder.

A comprehensive learning resource and infrastructure guide covering the entire AI technical stack, from hardware architecture and GPU programming to cloud-native deployment and large language model training.

Textus is a free, open‑source Markdown editor that adds AI‑driven writing help. It works on macOS, Windows and Linux, supports any OpenAI‑compatible LLM (including local Ollama models), and keeps your files and API keys strictly on your device.

A mobile and macOS application for running open-source LLMs and multimodal AI models fully on-device, providing a private and offline experience for chat, media search, and audio transcription.

RTranslator is a free, offline, real-time translation app for Android. It runs two AI models (Meta's NLLB for translation, OpenAI's Whisper for speech recognition) entirely on-device, so no internet is needed and privacy is preserved. It has three modes: Conversation (two phones connected via Bluetooth translate a live dialogue in both directions), WalkieTalkie (one phone, two people taking turns), and classic text translation. It requires at least 6GB of RAM, supports 31 languages at full quality plus more at reduced quality, and is still in beta with a known Bluetooth-drop bug. Version 3.0 is in development with better translation models and full open-source status.

A fork of ida‑pro‑mcp that provides a hardened, multi‑instance MCP server for headless IDA Pro, enabling AI agents (Claude Code, etc.) to drive reverse‑engineering tasks via a secure HTTP API.

An open-source PostgreSQL extension that enables vector similarity search, allowing users to store and query embeddings alongside relational data with ACID compliance.

An agent skill for AI coding assistants that provides specialized knowledge to help them write modern, performant, and accessible SwiftUI code while avoiding common LLM mistakes.

Stagehand is an SDK for browser agents that lets you control a browser with natural-language instructions—observe, act, and extract—without brittle CSS selectors. It works in TypeScript, Python, and Go, self-heals when sites change, and offers cloud integration with token-efficient caching.

A design system and AI skill for creating professional one- or two-ink editorial prints, posters, and packaging with a focus on mechanical reproduction aesthetics.

Agent Sandbox adds a Kubernetes CRD for a single, stateful, isolated pod (a “sandbox”). It’s designed for AI‑related workloads that need secure, long‑running containers with stable identity, persistent storage, and fast warm‑pool provisioning. Install via a single manifest, then create `Sandbox` or `SandboxClaim` resources; the controller creates pods using runtimes like gVisor or Kata. Go/Python SDKs and an optional router make programmatic use easy.

Claudex Loop is a multi-agent workflow that improves AI coding reliability by requiring independent AI models to review plans and inspect code generated by a primary builder.

ChatGPT2API is a reverse-engineered wrapper that exposes ChatGPT's official image generation and editing abilities as an OpenAI-compatible API, complete with account pool management, an online drawing workbench, and Docker-based self-hosting. It helps developers use ChatGPT image models without managing individual accounts manually.

Windmill is an open-source, self-hostable platform where you write scripts (Python, TypeScript, Go, Bash, etc.) and Windmill automatically turns them into UIs, APIs, cron jobs, and workflows. It supports AI-assisted building with review guardrails, managed data tables on Postgres, DuckDB/DuckLake for analytics, and claims to be the fastest self-hostable workflow engine. Backend is Rust + Postgres, frontend is Svelte 5, sandboxing via nsjail. Self-host via Docker or Kubernetes. AGPLv3 license with commercial options.

Ncurator is an AI-powered desktop assistant that automates office document management, local file editing, and task execution through a secure local interface.

An open-source flight-control firmware for ESP32-family microcontrollers (ESP32, ESP32-S3, ESP32-C3) that flies small drones like the Chenguang E1 series. It handles attitude sensing, motor PWM control, low-battery safety, LED status, and a browser-based remote control — all configurable through a debug console.

Free-LLM is a community-maintained, MIT-licensed directory of 270+ free LLM models from 51 providers plus 9 local/self-hosted tools. It offers credit-card transparency, ready-to-run code snippets, and a comparison tool, all backed by the live site free-llm.com. Providers are categorized into permanent free tiers, renewable credits, one-time trials, and self-hosted options. Most providers use OpenAI-compatible endpoints, making integration straightforward.

UFO is an AI agent framework that enables complex automation across multiple devices and operating systems, utilizing a DAG-based orchestration system to coordinate tasks between Windows, Linux, and Android.

botmux is a self‑contained binary that lets you start and control AI‑coding CLI agents (like Claude‑Code, Codex, Gemini, etc.) from Feishu/Lark messages. Each chat message spawns an isolated CLI process; its output streams back as live‑updating cards, and you can interact via a web terminal, hand‑off sessions, schedule tasks, or run multiple bots in the same group. Installation is a one‑line script (or npm) and works on Linux/macOS (WSL2 for Windows).

A versatile web-access toolkit for Pi agents that provides unified search, content extraction, and video analysis with intelligent fallback routing across dozens of providers.

TECHO5 is open firmware that replaces Amazon's Fire OS and Alexa on Echo Show 5, Show 8, and other devices, running a small Alpine Linux image with a Go daemon to make them fast, private Home Assistant voice assistants with touch screens.

A plain-English dictionary that translates AI coding jargon into clear, actionable concepts to help developers understand how models work and why they behave the way they do.

A comprehensive open-source tool that automates video translation by integrating speech recognition, LLM-based translation, and AI dubbing into a single workflow.

TeamAI‑CLI is an npm‑distributed command‑line tool that lets a software team store and automatically sync all AI‑related assets—prompt skills, safety rules, environment configs, model references, and usage logs—in a shared Git repository. After a one‑time `teamai init <repo‑url>` setup, every AI session (Claude, Codex, Cursor, Copilot, etc.) pulls the latest team assets, ensuring consistent, version‑controlled AI behavior across the whole team.

A high-performance audio processing library for Apple Silicon that enables local inference for text-to-speech, speech-to-text, and music generation models using the MLX framework.

X‑AnyLabeling is a cross‑platform desktop application for AI‑assisted annotation of images, video, text, and 3D point clouds. It bundles dozens of state‑of‑the‑art vision models (YOLO, SAM, Grounding‑DINO, etc.) and supports local or remote inference, offering tools for detection, segmentation, OCR, pose, VQA, and more, with export to common dataset formats.

Ragent AI is a production‑ready Java platform for building Agentic RAG applications. It provides hybrid retrieval (vector, keyword, graph, web), advanced query understanding, model tiering with fault‑tolerance, multi‑level session memory, Redis‑based traffic protection, and a full ingestion‑to‑answer loop. The code is split into seven Maven modules (infra‑ai, system, rag, agent, bootstrap, framework, mcp‑server) that isolate AI vendor, storage, and retrieval concerns, making components interchangeable. v2 uses a ReAct‑style agent loop with MCP tool calls, Skills, and human‑in‑the‑loop confirmations. Production features include circuit‑breaker, transactional RocketMQ, idempotency, tracing (Langfuse optional), SSE streaming, and Sa‑Token security. A React UI offers both chat with citation/agent view and an admin console for knowledge‑base and intent‑tree management. Targeted at Java backend engineers and interview‑preparation candidates who need a real‑world, extensible AI stack rather than a toy demo.

An AI agent for the partnership card game GuanDan that uses structure-aware retrieval and PPO reinforcement learning to optimize long-term strategic play.

Ghostex is a cross-platform workspace for managing multiple AI coding agents, providing a unified interface for chat, terminal access, and IDE integration that persists across sessions and devices.

The project is a curated, community-maintained catalog of plugins, skills, MCP servers, and agent tools for AI assistants. It lists extensions for many popular AI coding tools, provides trust scores based on security scanning, and offers both human-readable and machine-readable formats. It is a discovery resource, not an installer.

An open-source platform for building enterprise-grade AI agents that integrates RAG pipelines and agentic workflows to create intelligent knowledge bases and customer service tools.

Rig is a Rust library and terminal‑based coding agent for building LLM applications. It offers a plugin‑centric architecture, support for 20+ model providers and 10+ vector‑store back‑ends, memory management, OpenTelemetry tracing, and a self‑reloading Bevy‑powered agent that can be scripted via JSON.

A data pipeline and transformation suite that uses knowledge graphs to extract structured data from unstructured text, enhancing an LLM's ability to reason over private data.

EditaPlot is a Codex-powered skill that turns tabular data into editable Origin scientific plots (OPJU, PNG, PDF, TIF). It first confirms how each column is used, then calls your local Origin to generate the plot — never inventing values and always producing editable results.

A Kubernetes-native virtualization middleware that enables GPU sharing, resource isolation, and device-aware scheduling across heterogeneous AI accelerators.

An open-source agent operating system that uses isolated WorkSpaces, white-box memory, and smart model routing to manage long-running, multi-project AI productivity tasks.

A tool to extract design systems from any website into design tokens and enforce visual consistency in CI/CD to prevent design drift.

A desktop-based AI stock analysis tool that integrates LLM agents with financial data tools to provide automated market research, technical analysis, and strategy backtesting.

ATAS is a research framework that generates videos from a screenplay while maintaining a persistent 3‑D world state, enabling consistent re‑filming of past moments from new angles.

FunASR is an industrial-grade speech recognition toolkit from Alibaba's ModelScope team. It provides a modular Python API, CLI, and server for offline/streaming ASR, plus VAD, punctuation, speaker diarization, emotion recognition, and edge deployment via llama.cpp. It includes multiple pre-trained models (e.g., Fun-ASR-Nano, SenseVoiceSmall) and supports GPU, CPU, and edge devices.

Maxun is an open‑source platform for turning any website into a structured API. It provides web‑scraping, crawling, monitoring, and AI‑assisted data extraction through a visual dashboard, REST API, SDKs, and CLI, and can be self‑hosted via Docker or run as a hosted service.

An open protocol that enables communication and interoperability between AI agents built on different frameworks and servers, allowing them to collaborate without exposing internal logic.

FinRobot is an open‑source, multi‑agent AI platform for high‑stakes financial analysis. It separates deterministic Python‑based valuation (DCF, LBO, etc.) from LLM‑generated narration, orchestrates specialized agents (data, analysis, modeling, synthesis, reporting, plus debate agents), and ships three versions: a production desktop app (V2), a self‑hosted web report generator (V1), and an educational AutoGen library (V0). The system integrates multiple market data sources, provides 56 analyst playbooks, and can be run locally via a CLI, a browser UI, or a native Tauri desktop client.